CROSS-REFERENCEThis patent application is a continuation of, and claims priority to, U.S. patent application Ser. No. 11/319,964 (now U.S. Pat. No. 8,537,761), filed on Dec. 28, 2005, and entitled “INCORPORATION OF MESH BASE STATIONS IN A WIRELESS SYSTEM.” The entirety of the aforementioned application is incorporated by reference herein.
BACKGROUNDFIG. 1 shows ageneric mesh network100 according to prior art. Generic wireless mesh network100 (also referred as a client mesh network), includes nodes (subscriber stations)101 and117 in a wireless network forwarding traffic cooperatively over multiple radio links. Some of the participating nodes have wired connectivity to the Internet and hence serve as gateway nodes providing internet connectivity to the entire network. The architecture is economical when coverage, and not necessarily capacity, of the network is of primary concern.Mesh network100 exemplifies a typical mesh network. Current commercial, community, and public safety mesh networks are typically compatible with WiFi®, which is based on the IEEE 802.11 standard. Commercial players include companies such as Motorola, Nokia, Microsoft, Tropos, Mesh Networks, BelAir, Nortel, FireTide, Propagate, Strix, Mesh Dynamics, MeshAP, MIT Rooftop, Rice TAPs. Examples of municipality WiFi mesh networks can be found in Urbana, Kingsbridge, Queensland, MuniWireless (France), Philadelphia, San Francisco, Seattle, Portland, Chicago. The proliferation of mesh networks has already started even though the current WiFi-based trends promise coverage rather than capacity.
With the ubiquity of wireless subscriber stations, there is a real need in the market place to expand wireless mesh architecture to a wireless system to increase the coverage and the traffic capacity of a wireless system.
BRIEF DESCRIPTION OF THE DRAWINGSFIG. 1 shows a generic mesh network according to prior art.
FIG. 2 shows a mesh network according to various aspects.
FIG. 3 shows sector frequency allocations of a wireless base station according to prior art.
FIG. 4 shows a sector configuration of a wireless network according to prior art.
FIG. 5 shows a sector configuration of a wireless network according to an embodiment.
FIG. 6 shows simultaneous scheduling in a sector of a wireless network according to an embodiment.
FIG. 7 shows an adjacent sector configuration of a wireless network according to an embodiment.
FIG. 8 shows an example downlink scheduling in a sector according to an embodiment.
FIG. 9 shows examples of a modulation configurations in accordance with various embodiments.
FIG. 10 shows an example of simulation parameters in accordance with an embodiment.
FIG. 11 shows a directional antenna pattern in accordance with an embodiment.
FIG. 12 shows a multicell configuration in accordance with prior art.
FIG. 13 shows a multicell configuration in accordance with an embodiment.
FIG. 14 shows a scatter plot without a log normal distribution in accordance with prior art.
FIG. 15 shows a scatter plot with a log normal distribution in accordance with prior art.
FIG. 16 shows a downlink scatter plot for simultaneous scheduling without a log normal distribution in accordance with an embodiment.
FIG. 17 shows a downlink scatter plot for simultaneous scheduling with a log normal distribution in accordance with an embodiment.
FIG. 18 shows a downlink scatter plot for dedicated scheduling without a log normal distribution in accordance with an embodiment.
FIG. 19 shows a downlink scatter plot for dedicated scheduling with a log normal distribution in accordance with an embodiment.
FIG. 20 shows a cumulative density function of data rates for wireless subscriber stations in accordance with an embodiment.
FIG. 21 shows an outage rate for wireless subscriber stations in accordance with an embodiment.
FIG. 22 shows a sector throughput for a wireless system in accordance with an embodiment.
DETAILED DESCRIPTIONIn the following description of the various embodiments, reference is made to the accompanying drawings which form a part hereof, and in which is shown by way of illustration various embodiments can be practiced. It is to be understood that other embodiments can be utilized and structural and functional modifications can be made without departing from the scope of various aspects described herein. Definitions for the following terms are included to facilitate an understanding of the detailed description.
- mesh base station—an infrastructure entity that wirelessly relays data to and from a corresponding wireless base station;
- wireless resource allocation—a configurable wireless communication characteristic, e.g., frequency allocation (frequency division multiple access), code sequence allocation (code division multiple access), time slot allocation (time division multiple access), etc. Moreover, a wireless resource allocation can be specified as a combination of component resource allocations, e.g., a combination of frequency and time slot allocations.
 
In accordance with various embodiments, a mesh base station can utilize common wireless resource allocations as a corresponding wireless base station. In an aspect, a wireless base station and a mesh base station transmit to corresponding wireless subscriber stations during the same time period. As an example, regarding a scheduling downlink, mapping can be compatible with a frame structure that is supported by the IEEE 802.16 standard. Furthermore, either the wireless base station or the mesh base station can transmit to another wireless subscriber station during another time period of the frame structure.
In another aspect, a mesh base station obtains a data packet from a wireless base station over a backhaul link: that corresponds to a scheduled time period. The mesh base station consequently transmits the data packet to the designated wireless subscriber station during another scheduled time period. The other scheduled time period can be a dedicated time period or a simultaneous time period.
In yet another aspect, a wireless base station and a mesh base station receives data packets from corresponding wireless subscriber stations during the same time period. Data packets can also be transmitted to the wireless subscriber stations either in a symmetric manner or an asymmetric manner.
In one aspect, a mesh base station can be reconfigured to support wireless traffic if a wireless base station goes out of service or exceeds a predetermined traffic load. In such scenarios, a backhaul link is established to another wireless base station.
In another aspect, a wireless network is configured with two mesh base stations at an approximate boundary of two adjacent sector coverage areas. Each sector coverage area is supported by a corresponding wireless base station. Each mesh base station supports wireless subscriber stations within a configured coverage radius and connects to one of the wireless base stations over a backhaul link.
FIG. 2 shows a
mesh network200 according to embodiments. Mesh
network200 can be referred as an infrastructure mesh network because mesh base stations
211-
213 are considered part of the wireless network with
wireless base station209. (In contrast,
generic mesh network100 comprises only wireless subscriber stations to support a wireless mesh network.)
Mesh network200 is fundamentally different from
generic mesh network100 because of the fact that additional mesh base stations (mBS)
211-
213 are strategically deployed and controlled by
wireless system200 to wirelessly forward traffic from subscriber stations (SS)
203 and
207 to base station (BS)
209. The BS
mBS links
251 and
253 are referred as backhaul links. Additionally,
SS201 and
SS205 communicate directly with
BS209.
 FIG. 2 depicts single cell inmesh network200. A cell is defined as the area that aroundBS209 such that any SS in the coverage area can connect to the Internet via the particular BS. Of course, cells can overlap thus allowing users to select among multiple possible base stations.Infrastructure mesh network200 attempts to change the economics of micro-cell wireless networks by aggregating traffic for wired backhaul. Transmission cost for wired backhaul can be a significant cost in high capacity radio networks.Infrastructure mesh network200 also can provide advantages over a client mesh network, e.g.,generic mesh network100. For example, security, predictability, and manageability can be facilitated since the mesh base stations211-213 are centrally deployed and controlled. Unlike in a client mesh network, users do not forward (relay) any data packets. A user either directly communicates withBS209 or communicates withmBS211 or213, which then subsequently forwards traffic to and from theBS209. (Each communications link, as shown inFIG. 2 is bidirectional. The wireless subscriber station transmits to the wireless infrastructure on the uplink and receives from the wireless infrastructure on the downlink.)
Since
mBS211 or
213 is deployed by the
wireless system200, the mBS antenna is better placed than the antenna of a normal user and hence the BS
BS link is optimized to be a high-rate link. Moreover, mBS
SS links are typically shorter on average than mBS
BS links, thus requiring lesser transmission power and hence causing lesser interference to other users as well as other cells. Due to these lucrative advantages offered by an infrastructure mesh, most commercial WiFi mesh systems mentioned earlier are actually infrastructure mesh systems, i.e., the wireless system carefully deploys and controls the base stations.
According to embodiments, variations of
infrastructure mesh network200 can be supported. For example the BS
mBS backhaul251 or
253 can use a different spectrum than the spectrum used for the SS
BS and the SS
mBS links. This can lead to an economic problem due to the requirement of extra spectrum. Moreover, once a particular spectrum is dedicated to backhaul link
251 or
253, the corresponding frequency spectrum cannot be used for other purposes, thus preventing flexible use of the spectrum. As a result, the economical choice corresponds to having backhaul link
251 or
253 share the same spectrum as that being used by the clients. This choice has several other advantages as well such as requiring the same type of radio technology on the wireless subscriber stations (clients) the BS, and the mBS. Since the spectrum is shared, there is a natural tendency to design for flexible and frugal use of the shared spectrum. Owing to these reasons, one chooses to analyze the later option, i.e., to analyze, in terms of capacity and outage, the performance of an infrastructure mesh with a single spectrum being shared by all links in a cell.
According to an embodiment, in order to proceed with an analysis, one can impose simplifying assumptions. First, one assumes the existence of a centralized medium access control (MAC)packet radio system2 such as found in IEEE 802.16/WiMAX, CDMA EV-DO, UMTS-HSDPA, etc. (For example,BS209,mBS211, andmBS213 control access while wireless subscriber stations201-207 do not.) Second, one assumes that users are uniformly distributed in the analyzed area and that users always have backlogged data to send and receive, i.e., users have infinite bandwidth requirement. Finally, as a design principle, one imposes that the mBS support an infrastructure mesh need having “low complexity.” This design principle has the following consequences:
- the mBS should be smaller than the BS.
- the mBS should have a single radio to communicate with both the BS and subscribers; and
- the mBS should use an omni-directional antenna to communicate with the subscribers.
 
Consequently, one requires that an mBS (211,213) should have similar complexity as a SS (201-207), resulting in an mBS being as economical as wireless subscriber station. There can be some differences since mBS (211,213) can require a directional antenna forbackhaul link251 or253. However, mBS (211,213) typically is able to use the same radio for communicating through either the omni-directional antenna or the directional antenna through simple switches.
FIG. 3 shows sector frequency allocations for a wireless base station according to prior art.FIG. 3 shows a conventional cell with six sectors301-311, each using a different, non-interfering frequency spectrum. The BS is located at an approximate center of the cell and is assumed to have six different radios and correspondingly, six different directional antennas.
FIG. 4 shows asector coverage area401 of a wireless network that is supported by wireless base station according to prior art.Sector coverage area401 corresponds to any of the six sector areas301-311 as shown inFIG. 3.
FIG. 5 shows asector coverage area501 that is supported bywireless base station503 according to an embodiment.Mesh base stations505 and507 are incorporated at an approximate boundary ofsector coverage area501. Consequently,mesh base stations505 and507 are farthest fromwireless base station503 in a region wherebase station503 typically has the worst transmission characteristics. Moreover, wireless subscriber stations (nodes, not shown) at the extremities in a conventional sector need to transmit at a higher power, thus causing higher interference. The placement of a mesh base station at the base of the triangle representing a sector helps alleviate both issues. Instead of placing just one mesh base station at the middle of the base of the triangle, an embodiment places two mBS's505 and507 symmetrically in order to cover the entire base of the triangle. A single mBS can require the antenna pattern of the mBS to be contorted, resulting in complex antennas that are costly to build. Given the corresponding model, one wishes to analyze whether simultaneous use of spectrum by two different subscriber stations is feasible. If indeed it is possible to simultaneously schedule two users, one towireless base station503 and the other to one of themesh base stations505 or507, then one expects a throughput gain for the sector.
FIG. 6 shows simultaneous scheduling insector coverage area501 of a wireless network according to an embodiment. For example, wireless subscriber station (node A)601 can be scheduled to receive fromBS503 at the same time period a wireless subscriber station (node B)603 or wireless subscriber station (node C)605 is scheduled to receive frommBS505. This is because the interference from themBS505 atnode A601 is sufficiently attenuated and hence the Signal-to-Noise-and-InterferenceRatio (SINR) atnode A601 is sufficient for correct reception fromBS503. Similarly, the SINR atnode B603, even in the presence of interference fromBS503, is sufficient for satisfactory reception frommBS505.Node C605, even though not located insector coverage area501, can be supported bymBS505 becausenode C605 is within a coverage radius ofmBS505.
FIG. 7 shows adjacentsector coverage areas501 and701 of a wireless network according to an embodiment. Referring toFIG. 6, one can selectnode C605 rather thannode B603 to be scheduled simultaneously withnode A601. The SINR atnode C605 is typically higher than the SINR atnode B603 sinceBS503 is farther away fromnode C605 thanBS503 is fromnode B603. As a result, instead of analyzing just the single sector, one is motivated to analyzing a coverage area comprising two adjacent sector coverage areas as shown inFIG. 7.
Adjacentsector coverage areas501 and701 are assigned different frequency spectra.Mesh base station505 communicates withBS503 overbackhaul link751 even though many of the served users (e.g., node A705) can actually be located in the lower sector (sector coverage area701). Similarly, themBS707 can serve users (e.g., node B707) which are located in the upper sector (sector coverage area501) even thoughmBS707 communicates withBS703 overbackhaul link753.
FIG. 8 shows an example downlink scheduling insector coverage area501 according to an embodiment. A pair of wireless subscriber stations (e.g.,node B805 and node A803) is identified for simultaneous scheduling of communications withBS503 andmBS505.FIG. 8 shows an exemplary downlink scheduling frame (comprising data packets851-859) which is compatible with the frame structure currently used in IEEE 802.16. Nodes A andB803 and805 are simultaneously scheduled to receive fromBS503 andmBS505, respectively. Nodes C andD801 and807 are dedicatedly scheduled to receive fromBS503 andmBS859, respectively. The downlink frame contains the following periods:
- Backhaul period (corresponding to data packet851): In this period the data is forwarded from the BS to the mBS using the backhaul link. This data is meant for node Band D;
- Simultaneous Schedule period (corresponding todata packets853 and855): In this period the BS transmits data to node A and at the same time the mBS forwards the data meant for B that the mBS received in the immediately preceding backhaul period; and
- Dedicated Schedule period (corresponding todata packets857 and859): This period consists of two consecutive periods. In the first part the BS transmits data to node C and in the second part the mBS forwards data to node D.
 
While the above example illustrates transmission on the downlink (from the infrastructure to the wireless subscriber station), transmission can be scheduled on the uplink (from the wireless subscriber station to the infrastructure).
FIG. 9 shows an example of a
modulation scheme900 in accordance with an embodiment.
FIG. 9 shows different modulation schemes that can be used and the corresponding SINR thresholds and transmission rates, corresponding to different modulation configurations. In an embodiment, modulation scheme
905 (64QAM) is used for the BS
mBS backhaul link (e.g.,
links751 and
753 as shown in
FIG. 7).
 FIG. 10 shows an example of simulation parameters in accordance with an embodiment. The remaining parameters used for the simulation are shown inFIG. 10. In an embodiment, the Erceg-Greenstein model is used as the large scale fading model and most of the results are shown for terrain A of the Erceg-Greenstein model. Terrain A is representative of areas with moderate to heavy tree density. Simulations with the other terrain types were performed with similar results.
FIG. 11 shows a directional antenna pattern used for the antenna atwireless base stations503 and703 (as shown inFIG. 7) in accordance with an embodiment.Mesh base stations505 and507 and wireless subscriber stations705-707 utilize omni directional antennas. The directional antenna at BS (503,703) has a beamwidth of 30 degrees. The gain at an angle of departure of 0 degrees from the azimuth of the directional antenna, i.e., the peak gain of the antenna, is 0 dBi and the gain at an angle of departure of 30 degrees is −12 dBi. Since one assumes an antenna gain of 20 dBi, the corresponding effective gains are 20 dBi and 8 dBi, respectively.
FIG. 12 shows amulticell configuration1200 in accordance with prior art. In the scatter plots shown inFIGS. 14 and 15,sector coverage areas1201 and1203 are assigned the same frequency spectrum. In the analysis of the downlink throughput and coverage, interference from cells which are at most two cells away from the analyzed sector is considered.
FIG. 13 shows a
multicell configuration1300 in accordance with an embodiment. In the scatter plots shown in
FIGS. 16-19, downlink performance is determined in the presence of
mesh base stations1305 and
1307 serving
sector coverage areas1301 and
1303. In the analysis, the wireless system has a (1,6,6) spectrum configuration (corresponding each cell being assigned the same frequency spectrum, each cell being divided into
6 sectors, and each sector having a corresponding frequency allocation.) As will be discussed in
FIGS. 16-19, the improvement in the mesh system in terms of coverage and throughput more than compensates for the radio resources that are diverted towards the mBS
BS backhaul. The mesh base station is designed to be simple and hence cost effective, thus not creating any economic hurdle in their deployment. Simulations have shown a coverage improvement of around 80% whereas the sector throughput increases from 16 Mbps to 21 Mbps. In addition, embodiments can support power control so that transmissions are made at the optimum power required for a particular SINR to be achieved between the transmitter and the receiver. One consequently expects a decrease of inter-cell interference caused by wireless subscriber stations at the periphery of a cell.
 Embodiments also support uplink scheduling, which can be symmetric or asymmetric with respect to downlink scheduling. For example, a wireless subscriber station can communicate during different time periods for the uplink and the downlink. Also, a wireless subscriber station can communicate with a mesh base station in one direction and directly communicate with a wireless base station in the other direction.
FIG. 14 shows ascatter plot1400 without a log normal distribution in accordance with prior art.FIG. 15 shows ascatter plot1500 with a log normal distribution in accordance with prior art. (Scatter plots1400 and1500 correspond tomulti-cell configuration1200 as shown inFIG. 12.) Withscatter plot1400 no log normal variation was applied to the path loss, while withscatter plot1500 log normal variation was applied to the path loss.
FIGS. 16-19 show the points which can be simultaneously scheduled in the presence of mesh base stations and points which require dedicated scheduling.FIG. 16 shows adownlink scatter plot1600 for simultaneous scheduling without a log normal distribution in accordance with an embodiment.FIG. 17 shows adownlink scatter plot1700 for simultaneous scheduling with a log normal distribution in accordance with an embodiment.FIG. 18 shows adownlink scatter plot1800 for dedicated scheduling without a log normal distribution in accordance with an embodiment.FIG. 19 shows adownlink scatter plot1900 for dedicated scheduling with a log normal distribution in accordance with an embodiment. (Scatter plots1600,1700,1800, and1900 correspond tomulti-cell configuration1300 as shown inFIG. 1300.)FIGS. 16 and 18 show the scatter plots when no log normal variation is used to calculate path loss.FIGS. 17 and 19 show scatter plots in which log normal variation in path loss is included. (One expects that a log normal variation in the path loss provides results that better approximate an actual situation.) As a result there is no clear demarcation between the region which requires simultaneous scheduling and the region which requires dedicated scheduling as shown inFIGS. 17 and 19. However, this demarcation is clearly visible inFIGS. 16 and 18.
FIG. 20 shows a cumulative density function (CDF)2000 of data rates for wireless subscriber stations in accordance with an embodiment. Even though the actual transmission rates are as shown inFIG. 9,CDF plot2000 for the data rates in the presence of mesh base stations has transitions that do not match with the transmission rates shown inFIG. 9. This observation results when a wireless subscriber station communicates with a wireless base station via mesh base station, in which the actual data rate that the wireless subscriber station encounters is less than the transmission rate that the wireless subscriber station is able to transmit or receive at. Correspondingly, data must be transmitted over the backhaul link, during which time no other transmission can be performed in the sector (e.g., the scheduling as shown inFIG. 8).FIG. 20 suggests that a large number of wireless subscriber stations that were not able to communicate in the conventional case (corresponding tomulti-cell configuration1200 as shown inFIG. 12) are able to do so in the presence of mesh base stations (corresponding tomulti-cell configuration1300 as shown inFIG. 13).
FIG. 21 shows an outage rate for wireless subscriber stations using a QPSK ½ modulation scheme in accordance with an embodiment. One observes a significant decrease in the outage rate comparing a configuration without mesh base stations (rates2101 and2105) to a configuration with mesh base stations (rates2103 and2107).FIG. 22 shows a corresponding sector throughput for a wireless system in accordance with an embodiment. One observes an increase of the sector throughput comparing a configuration without mesh base stations (data throughputs2201 and2205) to a configuration with mesh base stations (data throughputs2203 and2207).
Embodiments support operational scenarios in which a wireless base station goes out of service. As an example, refer to the wireless network as shown inFIG. 7. As previously discussed,mesh base station505 communicates withwireless base station503 overbackhaul link751, andmesh base station507 communicates withwireless base station703 overbackhaul link753. Ifwireless base station503 goes out of service, then meshbase station505 establishes a backhaul link towireless base station703 and can also expand its coverage intosector coverage area501. (The backhaul link can be established in a number of ways. For example, a directional communication path can be established betweenmesh base station505 andwireless base station703. Alternatively, another time period can be scheduled for backhauling betweenmesh base station505 andwireless base station703.) Traffic for wireless subscriber stations within the coverage radius ofmesh base station505 are consequently diverted towireless base station703.
Embodiments also support operational scenarios in which a wireless base station exceeds a predetermined level of traffic (i.e., overload). As an example, refer to the wireless network as shown inFIG. 7. As previously discussed,mesh base station505 communicates withwireless base station503 overbackhaul link751, andmesh base station507 communicates withwireless base station703 overbackhaul link753. Ifwireless base station503 exceeds a predetermined traffic limit, then meshbase station505 establishes a backhaul link towireless base station703 so that traffic can be divertedwireless base station503. In such a scenario, traffic for wireless subscriber stations within the coverage radius ofmesh base station505 is consequently supported bywireless base station703. Moreover, the coverage radius ofmesh base station505 can be adjusted to change the number of wireless subscriber stations that are diverted fromwireless base station503 towireless base station703. The coverage area ofmesh base station505 can be adjusted by adjusting the transmit power level and/or receive sensitivity.
As can be appreciated by one skilled in the art, a computer system with an associated computer-readable medium containing instructions for controlling the computer system can be utilized to implement the exemplary embodiments that are disclosed herein. The computer system can include at least one computer such as a microprocessor, a digital signal processor, and associated peripheral electronic circuitry. Other hardware approaches such as utilizing a digital signal processor (DSP), utilizing a field programmable gate array (FPGA), etc. can also be used to implement the exemplary embodiments.
Those skilled in the art will appreciate that there are numerous variations and permutations of the above described systems and techniques that fall within the spirit and scope of embodiments disclosed herein as set forth in the appended claims.